cleveragents-core/tools/controller/master/forgejo_http.py

"""HTTP adapters wiring the controller's callback protocols to
``tools/_claim_runtime``.

The discovery / scheduler-prefetch / forgejo_writes / merging modules
all take callback functions for Forgejo I/O. Tests inject synthetic
callbacks; production wires them via this module's
``build_callbacks(cfg)`` factory.

The factory takes a ``RuntimeContext`` (from ``_claim_runtime``) and
returns a frozen ``ForgejoCallbacks`` dataclass with every callback
the controller needs. The callbacks are thin closures over
``_claim_runtime.get / post / patch / delete``.

Path conventions match Forgejo's API:
- list PRs:       GET /repos/{owner}/{repo}/pulls?state=open
- list issues:    GET /repos/{owner}/{repo}/issues?state=open&type=issues
- list comments:  GET /repos/{owner}/{repo}/issues/{n}/comments
- post comment:   POST /repos/{owner}/{repo}/issues/{n}/comments
- get labels:     GET /repos/{owner}/{repo}/issues/{n}/labels
- add label:      POST /repos/{owner}/{repo}/issues/{n}/labels
- remove label:   DELETE /repos/{owner}/{repo}/issues/{n}/labels/{label_id}
- merge PR:       POST /repos/{owner}/{repo}/pulls/{n}/merge

Forgejo treats issues + PRs interchangeably at the timeline level
(labels + comments use issue endpoints for both).
"""
from __future__ import annotations

import logging
from dataclasses import dataclass
from typing import Any, Protocol

from . import forgejo_writes as fw
from . import merging as mg
from . import discovery as ds
from . import prefetch as pf
from . import reconciliation as rec

logger = logging.getLogger(__name__)


class _ClaimRuntime(Protocol):
    """Just the parts of ``_claim_runtime`` we use. Lets tests inject
    a fake without importing the real module."""

    def get(self, path: str, cfg: Any) -> dict[str, Any]: ...
    def post(self, path: str, cfg: Any, body: Any) -> dict[str, Any]: ...
    def delete(self, path: str, cfg: Any) -> dict[str, Any]: ...


# get_ci_status(owner, repo, head_sha) → Forgejo combined-status dict
# (``{"state": "...", "statuses": [...]}``) or None on fetch failure.
from collections.abc import Callable as _Callable
GetCIStatusCallback = _Callable[[str, str, str], dict | None]


@dataclass
class ForgejoCallbacks:
    """Bundle of every callback the controller needs. Built by
    ``build_callbacks``. Each callback is a thin closure over
    ``_claim_runtime``."""

    list_prs: ds.ListPRsCallback
    list_issues: ds.ListIssuesCallback
    list_comments: fw.ListCommentsCallback
    post_comment: fw.PostCommentCallback
    get_labels: fw.GetLabelsCallback
    add_label: fw.AddLabelCallback
    remove_label: fw.RemoveLabelCallback
    merge_pr: mg.MergeCallback
    # Reconciliation callbacks (Phase 1g):
    get_pr_state: rec.GetPRStateCallback
    get_issue_state: rec.GetIssueStateCallback
    # CI status callback (Phase 1k++++ trial — wired to ci_status_poll):
    get_ci_status: "GetCIStatusCallback"
    # Prefetch callbacks (Phase 1h):
    get_pr_details: pf.GetPRDetailsCallback
    get_pr_diff: pf.GetPRDiffCallback
    list_pr_reviews: pf.ListPRReviewsCallback
    list_pr_comments: pf.ListPRCommentsCallback


def build_callbacks(
    cfg: Any, *, runtime: _ClaimRuntime | None = None,
) -> ForgejoCallbacks:
    """Wire callbacks for a given RuntimeContext.

    Tests pass a fake ``runtime`` with ``.get / .post / .delete``
    methods that return the standard ``{"status": int, "body": ...}``
    shape. Production omits ``runtime`` and the wiring uses the real
    module.
    """
    if runtime is None:
        import sys
        from pathlib import Path
        tools_dir = Path(__file__).resolve().parents[2]
        if str(tools_dir) not in sys.path:
            sys.path.insert(0, str(tools_dir))
        from tools import _claim_runtime as runtime  # type: ignore[no-redef]

    return ForgejoCallbacks(
        list_prs=_make_list_prs(cfg, runtime),
        list_issues=_make_list_issues(cfg, runtime),
        list_comments=_make_list_comments(cfg, runtime),
        post_comment=_make_post_comment(cfg, runtime),
        get_labels=_make_get_labels(cfg, runtime),
        add_label=_make_add_label(cfg, runtime),
        remove_label=_make_remove_label(cfg, runtime),
        merge_pr=_make_merge_pr(cfg, runtime),
        get_pr_state=_make_get_pr_state(cfg, runtime),
        get_issue_state=_make_get_issue_state(cfg, runtime),
        get_pr_details=_make_get_pr_details(cfg, runtime),
        get_pr_diff=_make_get_pr_diff(cfg, runtime),
        list_pr_reviews=_make_list_pr_reviews(cfg, runtime),
        list_pr_comments=_make_list_pr_comments(cfg, runtime),
        get_ci_status=_make_get_ci_status(cfg, runtime),
    )


# ─── discovery ───────────────────────────────────────────────────────


def _make_list_prs(cfg, runtime):
    def list_prs(owner: str, repo: str) -> list[dict]:
        path = f"/repos/{owner}/{repo}/pulls?state=open"
        resp = runtime.get(path, cfg)
        return _list_or_empty(resp, "list_prs")
    return list_prs


def _make_list_issues(cfg, runtime):
    def list_issues(owner: str, repo: str) -> list[dict]:
        # Forgejo's /issues endpoint returns both issues + PRs by
        # default; type=issues filters out PRs (which list_prs handles).
        path = f"/repos/{owner}/{repo}/issues?state=open&type=issues"
        resp = runtime.get(path, cfg)
        return _list_or_empty(resp, "list_issues")
    return list_issues


# ─── status comments ─────────────────────────────────────────────────


def _make_list_comments(cfg, runtime):
    def list_comments(owner: str, repo: str, pr_number: int) -> list[dict]:
        path = f"/repos/{owner}/{repo}/issues/{int(pr_number)}/comments"
        resp = runtime.get(path, cfg)
        return _list_or_empty(resp, "list_comments")
    return list_comments


def _make_post_comment(cfg, runtime):
    def post_comment(
        owner: str, repo: str, pr_number: int, body: str,
    ) -> dict:
        path = f"/repos/{owner}/{repo}/issues/{int(pr_number)}/comments"
        resp = runtime.post(path, cfg, {"body": body})
        if int(resp.get("status") or 0) not in (200, 201):
            raise RuntimeError(
                f"post_comment HTTP {resp.get('status')}: "
                f"{_summarise_body(resp.get('body'))}"
            )
        body_obj = resp.get("body") or {}
        if not isinstance(body_obj, dict):
            return {"id": None}
        return body_obj
    return post_comment


# ─── labels ──────────────────────────────────────────────────────────


def _make_get_labels(cfg, runtime):
    def get_labels(owner: str, repo: str, pr_number: int) -> list[dict]:
        path = f"/repos/{owner}/{repo}/issues/{int(pr_number)}/labels"
        resp = runtime.get(path, cfg)
        return _list_or_empty(resp, "get_labels")
    return get_labels


def _make_add_label(cfg, runtime):
    def add_label(
        owner: str, repo: str, pr_number: int, label_name: str,
    ) -> bool:
        path = f"/repos/{owner}/{repo}/issues/{int(pr_number)}/labels"
        # Forgejo accepts {"labels": ["label_name"]} on POST.
        resp = runtime.post(path, cfg, {"labels": [label_name]})
        return int(resp.get("status") or 0) in (200, 201)
    return add_label


def _make_remove_label(cfg, runtime):
    def remove_label(
        owner: str, repo: str, pr_number: int, label_name: str,
    ) -> bool:
        # Forgejo's DELETE-by-name endpoint:
        #   DELETE /repos/{owner}/{repo}/issues/{n}/labels/{label_name}
        # (some Forgejo deployments use label_id; if so, the caller
        # would need to look it up first via get_labels)
        from urllib.parse import quote
        path = (
            f"/repos/{owner}/{repo}/issues/{int(pr_number)}/"
            f"labels/{quote(label_name, safe='')}"
        )
        resp = runtime.delete(path, cfg)
        return int(resp.get("status") or 0) in (200, 204, 404)
    return remove_label


# ─── merge ───────────────────────────────────────────────────────────


def _make_merge_pr(cfg, runtime):
    def merge_pr(owner: str, repo: str, pr_number: int) -> mg.MergeResponse:
        path = f"/repos/{owner}/{repo}/pulls/{int(pr_number)}/merge"
        try:
            resp = runtime.post(path, cfg, {"Do": "merge"})
        except Exception as exc:  # noqa: BLE001 — bubble as transient
            # Synthetic 503 — the merging handler treats this as a
            # retryable failure.
            return mg.MergeResponse(
                status_code=503,
                error_message=f"merge HTTP call raised: {exc}",
            )
        status = int(resp.get("status") or 0)
        body = resp.get("body")
        err = _summarise_body(body) if status >= 400 else None

        # For 404, try to fetch the PR state to distinguish
        # externally-merged vs externally-closed. If the PR fetch
        # itself fails (5xx / 404 / other), leave pr_state=None
        # so the merging handler defaults to ABANDONED conservatively.
        pr_state: str | None = None
        if status == 404:
            try:
                pr_resp = runtime.get(
                    f"/repos/{owner}/{repo}/pulls/{int(pr_number)}", cfg,
                )
                if int(pr_resp.get("status") or 0) == 200:
                    pr_body = pr_resp.get("body") or {}
                    if isinstance(pr_body, dict):
                        if pr_body.get("merged") is True:
                            pr_state = "merged"
                        elif pr_body.get("state") == "closed":
                            pr_state = "closed"
                        else:
                            pr_state = "open"
            except Exception:
                # Couldn't determine; leave None → handler defaults
                # to ABANDONED per its conservative policy.
                pass

        return mg.MergeResponse(
            status_code=status, pr_state=pr_state, error_message=err,
        )
    return merge_pr


# ─── helpers ─────────────────────────────────────────────────────────


def _list_or_empty(resp: dict, action: str) -> list[dict]:
    """Normalize a list-GET response into a list of dicts."""
    status = int(resp.get("status") or 0)
    if status != 200:
        logger.warning("%s: HTTP %s; treating as empty", action, status)
        return []
    body = resp.get("body")
    if not isinstance(body, list):
        return []
    return [item for item in body if isinstance(item, dict)]


def _summarise_body(body: Any) -> str:
    """Truncate a Forgejo error body to fit in a one-line error."""
    if body is None:
        return ""
    s = str(body)
    if len(s) <= 200:
        return s
    return s[:200] + f"… ({len(s)} chars)"


# ─── reconciliation (Phase 1g) ───────────────────────────────────────


def _make_get_pr_state(cfg, runtime):
    def get_pr_state(owner: str, repo: str, pr_number: int) -> dict | None:
        path = f"/repos/{owner}/{repo}/pulls/{int(pr_number)}"
        resp = runtime.get(path, cfg)
        status = int(resp.get("status") or 0)
        if status == 404:
            return None
        if status != 200:
            logger.warning(
                "get_pr_state: HTTP %s for %s/%s #%d",
                status, owner, repo, pr_number,
            )
            # Treat non-200/404 as "couldn't determine" — raise so the
            # reconciliation tick records fetch-failed (rather than
            # silently transitioning to STUCK).
            raise RuntimeError(f"get_pr_state HTTP {status}")
        body = resp.get("body")
        if not isinstance(body, dict):
            raise RuntimeError("get_pr_state: non-dict body")
        return body
    return get_pr_state


def _make_get_issue_state(cfg, runtime):
    def get_issue_state(owner: str, repo: str, issue_number: int) -> dict | None:
        path = f"/repos/{owner}/{repo}/issues/{int(issue_number)}"
        resp = runtime.get(path, cfg)
        status = int(resp.get("status") or 0)
        if status == 404:
            return None
        if status != 200:
            raise RuntimeError(f"get_issue_state HTTP {status}")
        body = resp.get("body")
        if not isinstance(body, dict):
            raise RuntimeError("get_issue_state: non-dict body")
        return body
    return get_issue_state


# ─── prefetch (Phase 1h) ─────────────────────────────────────────────


def _make_get_pr_details(cfg, runtime):
    def get_pr_details(owner: str, repo: str, pr_number: int) -> dict | None:
        path = f"/repos/{owner}/{repo}/pulls/{int(pr_number)}"
        try:
            resp = runtime.get(path, cfg)
        except Exception as exc:  # noqa: BLE001 — transport failure
            logger.warning(
                "get_pr_details: transport error for %s/%s #%d: %s",
                owner, repo, pr_number, exc,
            )
            return None
        status = int(resp.get("status") or 0)
        if status == 404:
            return None
        if status != 200:
            logger.warning(
                "get_pr_details: HTTP %s for %s/%s #%d",
                status, owner, repo, pr_number,
            )
            return None
        body = resp.get("body")
        return body if isinstance(body, dict) else None
    return get_pr_details


def _make_get_pr_diff(cfg, runtime):
    def get_pr_diff(owner: str, repo: str, pr_number: int) -> str | None:
        # Forgejo: GET /repos/{owner}/{repo}/pulls/{n}.diff returns
        # the raw unified diff as text/plain.
        path = f"/repos/{owner}/{repo}/pulls/{int(pr_number)}.diff"
        try:
            resp = runtime.get(path, cfg)
        except Exception as exc:  # noqa: BLE001
            logger.warning(
                "get_pr_diff: transport error for %s/%s #%d: %s",
                owner, repo, pr_number, exc,
            )
            return None
        status = int(resp.get("status") or 0)
        if status != 200:
            return None
        body = resp.get("body")
        if isinstance(body, str):
            return body
        if isinstance(body, (bytes, bytearray)):
            try:
                return bytes(body).decode("utf-8", errors="replace")
            except Exception:
                return None
        return None
    return get_pr_diff


def _make_list_pr_reviews(cfg, runtime):
    def list_pr_reviews(owner: str, repo: str, pr_number: int) -> list[dict]:
        path = f"/repos/{owner}/{repo}/pulls/{int(pr_number)}/reviews"
        resp = runtime.get(path, cfg)
        return _list_or_empty(resp, "list_pr_reviews")
    return list_pr_reviews


def _make_list_pr_comments(cfg, runtime):
    def list_pr_comments(owner: str, repo: str, pr_number: int) -> list[dict]:
        # Forgejo: PR comments use the issues endpoint (PRs ARE issues
        # at the comment/label level).
        path = f"/repos/{owner}/{repo}/issues/{int(pr_number)}/comments"
        resp = runtime.get(path, cfg)
        return _list_or_empty(resp, "list_pr_comments")
    return list_pr_comments


# ─── CI status (Phase 1k++++ trial — wired to ci_status_poll) ───────


def _make_get_ci_status(cfg, runtime):
    def get_ci_status(owner: str, repo: str, head_sha: str) -> dict | None:
        if not head_sha:
            return None
        path = f"/repos/{owner}/{repo}/commits/{head_sha}/status"
        try:
            resp = runtime.get(path, cfg)
        except Exception as exc:  # noqa: BLE001 — transient
            logger.warning(
                "get_ci_status: transport error for %s/%s @%s: %s",
                owner, repo, head_sha[:12], exc,
            )
            return None
        status = int(resp.get("status") or 0)
        if status != 200:
            logger.warning(
                "get_ci_status: HTTP %s for %s/%s @%s",
                status, owner, repo, head_sha[:12],
            )
            return None
        body = resp.get("body")
        return body if isinstance(body, dict) else None
    return get_ci_status


__all__ = ["ForgejoCallbacks", "GetCIStatusCallback", "build_callbacks"]